Variação espacial da biodiversidade arbórea Latino-Americana: padrões e determinantes

Abstract

Bioregionalization is the effort to identify and map geographical boundaries for distinct biota at different spatial scales. Clear delineation of biogeographic units based on community data can provide a framework for testing hypotheses and for directing biodiversity conservation efforts. The aim of this Thesis is to identify and map tree biogeographic units in Latin America at different spatial scales, investigate the floristic relationships between the identified units and the main factors responsible for the formation of floristic units in geographic space. The Thesis is structured in three articles. We used a community-level modelling approach, and hierarchical and non-hierarchical classification techniques to map the biogeographic units at different spatial scales. In Article I, we compiled woody species occurrence data from 301 assemblages distributed over the geographical extent of the Amazon rainforest. Occurrence data were used to model the B-diversity along the entire extent of the Amazon and thus to map floristic units (subregions). We used multinomial logistic regression model and statistical deviance partitioning to investigate the influence of environmental, historical and human factors on the distribution of Amazonian subregions. We identified 13 woody subregions in the entire Amazon Forest. The hierarchical subregion classification showed a broad Andean-Cratonic east–west division. Variation in subregions was explained jointly by human factors and spatial structure followed by environmental factors and spatial structure combined. In article II, we propose two bioregionalization schemes for tree species of the Cerrado-Pantanal complex, based on occurrence and abundance data. We also evaluated the contribution of three sets of determinants of the occurrence- and abundance-based subregions. For modelling, we compiled data on tree species composition from 894 local assemblages based on species occurrences, and from 658 local assemblages based on species abundances. We identified 18 occurrence- and four abundance-based subregions in the Cerrado-Pantanal complex. Variation in subregions were explained mainly by environmental factors and spatial structure in both occurrence and abundance datasets. The occurrence and abundance-based subregions are complementary approaches to disentangle the macroecological patterns and to conservation planning in the Cerrado-Pantanal complex. In Article III, we present regionalizations of the tree flora of Latin America using two dimensions of B-diversity, taxonomic and phylogenetic. We assessed taxonomic and phylogenetic floristic relationships between groups at different spatial scales. We used species occurrence data from 7306 tree assemblages to model taxonomic and phylogenetic B-diversity across Latin America. We identified 128 taxonomic tree subregions and 101 phylogenetic subregions. The tree subregions were grouped into 54 mesoregions, 16 macroregions and 9 supraregions. Phylogenetic subregions were grouped into 46 mesoregions and 15 macroregions. Our proposed bioregionalization in Latin America revealed a detailing of tree biodiversity patterns not usually presented in continental-scale studies, revealing ecotone areas between known biome boundaries, as well as the evolutionary and geographic history between regions.